Skip to main content

Installation Using uv

This guide covers how to install dtx using uv, a fast Python package installer and resolver.

uv is a great option if you:

  • Want faster installations (especially in CI/CD environments).
  • Are setting up fresh environments.
  • Want to use modern Python packaging tools.

Prerequisites​

  • Python >= 3.8
  • uv installer

Step 1: Install uv​

You can install uv globally with:

curl -LsSf https://astral.sh/uv/install.sh | sh

Verify the installation:

uv --version

Step 2: Install dtx Core​

To install dtx without local model dependencies:

uv pip install dtx

This installs the base CLI for:

  • Generating scopes
  • Generating plans
  • Running cloud-based models or using ddtx Docker CLI

If you want to run local models (e.g., Tiny-LLM, GPT-2, Hugging Face models), install with extras:

uv pip install dtx[torch]

This will install:

  • torch — Deep learning backend
  • transformers — Hugging Face model integration

Recommended if:

  • You want local/offline model execution
  • You want to integrate with Ollama or Hugging Face models

Step 4: Verify Installation​

Check if the CLI is available:

dtx --help

You should see a list of available commands.

Notes​

  • uv is optional but recommended for fast, reproducible installations.
  • No Docker is required if you install locally with uv.
  • For cloud models (OpenAI, Hugging Face), remember to configure your environment variables:

See: Environment Variables Setup